Information generation method, device and equipment, computer readable medium and program product

Abstract

The embodiment of the invention discloses an information generation method, device and equipment, a computer readable medium and a program product. A specific embodiment of the method comprises the steps of obtaining a user related portrait corresponding to a target user and historical order amount related information; according to the user related portrait and the historical order amount related information, generating a logistics prediction order amount in a target time period by using a pre-trained logistics order amount generation model; obtaining the actual logistics order amount of the target user in the target time period; and according to the actual logistics order amount and the logistics prediction order amount, order amount influence information in the target time period is generated. The implementation mode is related to artificial intelligence, and the influence condition of the first logistics platform on the order quantity change of the target user in the target time period can be accurately generated.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the field of computer technology, and in particular to information generation methods, devices, equipment, computer-readable media, and program products. Background Art

[0002] At present, in the logistics industry, when the logistics scale of users (i.e. customers) reaches a certain level, scientific management of the user's logistics life cycle plays an important role in improving the overall operational quality and health of the business. The improvement of single-user value uses technical means to identify changes in logistics orders, empower existing active users, and improve the user's single-user value through work methods such as visits. For the attribution analysis of changes in logistics orders, the methods usually used are: strategies based on data statistics, and methods based on regression and classification models to conduct attribution analysis of changes in logistics orders.

[0003] However, the inventors have found that when the above method is adopted, the following technical problems often occur:

[0004] Regarding the attribution analysis of changes in users' logistics order volume, existing models often fail to consider the impact of major influencing factors (for example, third-party factors) on the changes in logistics order volume, resulting in inaccurate attribution analysis of changes in logistics order volume.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the inventive concept and therefore it may contain information that does not form the prior art that is already known in this country to a person of ordinary skill in the art. Summary of the invention

[0006] The content of this disclosure is used to introduce concepts in a brief form, which will be described in detail in the detailed implementation section below. The content of this disclosure is not intended to identify the key features or essential features of the technical solution claimed for protection, nor is it intended to limit the scope of the technical solution claimed for protection.

[0007] Some embodiments of the present disclosure propose information generation methods, devices, equipment, computer-readable media, and program products to solve the technical problems mentioned in the above background technology section.

[0008] In a first aspect, some embodiments of the present disclosure provide an information generation method, including: obtaining user-related portraits and historical order quantity related information corresponding to a target user; based on the user-related portraits and the historical order quantity related information, using a pre-trained logistics order quantity generation model to generate a logistics forecast order quantity within a target time period, wherein the logistics order quantity generation model is trained based on a target training sample set, the target training samples are training samples whose first logistics platform does not affect the corresponding order quantity changes of users, and the logistics forecast order quantity is a representation of the forecast order quantity without being affected by the first logistics platform; obtaining the actual logistics order quantity of the target user within the target time period; based on the actual logistics order quantity and the logistics forecast order quantity, generating order quantity impact information within the target time period, wherein the order quantity impact information represents the order quantity changes under the influence of the first logistics platform.

[0009] Optionally, the user-related portraits include: a first historical user portrait corresponding to the target user, a second historical user portrait corresponding to a similar user, and the historical order quantity related information includes: a first historical order quantity change information corresponding to the target user, and a second historical order quantity change information corresponding to the similar user; and the logistics predicted order quantity within the target time period is generated by using a pre-trained logistics order quantity generation model based on the user-related portraits and the historical order quantity related information, including: obtaining logistics platform information corresponding to a second logistics platform set, wherein the second logistics platform is a logistics platform that has performed logistics processing for the target user; inputting the first historical user portrait, the second historical user portrait, the first historical order quantity change information, the second historical order quantity change information and the logistics platform information into the logistics order quantity generation model to generate the logistics predicted order quantity.

[0010] Optionally, the above method also includes: obtaining a weight set of order quantity status influencing factors corresponding to the order quantity status influencing factor set for the above target time period; obtaining a historical influencing factor data set corresponding to the above order quantity status influencing factor set; for each user order quantity change state in at least one user order quantity change state, executing the following first generation step: obtaining a user order quantity change information generation model corresponding to the above user order quantity change state; generating user order quantity change information using the above user order quantity change information generation model according to the above user-related portrait, the above historical influencing factor data set and the above order quantity status influencing factor weight set; in response to determining that the order quantity change situation represented by the above user order quantity change information is consistent with the order quantity change situation corresponding to the above user order quantity change state, determining at least one order quantity status influencing factor in the above order quantity status influencing factor set whose corresponding size of the order quantity status influencing factor weight is within the previous target number as at least one order quantity change attribution factor.

[0011] Optionally, the above-mentioned single quantity status influencing factor weight set is generated by the following steps: for each single quantity status influencing factor in the above-mentioned single quantity status influencing factor set, execute the following second generation step: obtain historical influencing factor data corresponding to the above-mentioned single quantity status influencing factor in a historical time period as target historical influencing factor data; input the above-mentioned target historical influencing factor data into the corresponding pre-trained time series neural network model to generate influencing factor feature information; input the above-mentioned influencing factor feature information, the first historical user portrait and the second historical user portrait into the pre-trained single quantity status influencing factor score generation model to generate the single quantity status influencing factor score; normalize the obtained single quantity status influencing factor score set to generate the above-mentioned single quantity status influencing factor weight set.

[0012] Optionally, the user order quantity change information generation model includes: a dense layer, an attention mechanism layer, an attribute feature information generation model and a user order quantity change information generation layer; and the user order quantity change information is generated by using the user order quantity change information generation model based on the user-related portrait, the historical influencing factor data set and the order quantity status influencing factor weight set, including: inputting the obtained influencing factor feature information set into the dense layer to generate an output information set; inputting the output information set and the order quantity status influencing factor weight set into the attention mechanism layer to generate an attention feature information set; obtaining the user attribute information set corresponding to the target user; using the attribute feature information generation model to generate attribute feature information for the user attribute information set; fusing the attribute feature information and the attention feature information set to generate fused feature information; inputting the fused feature information into the user order quantity change information generation layer to generate the user order quantity change information.

[0013] Optionally, the above-mentioned output information set and the above-mentioned single quantity state influencing factor weight set are input into the above-mentioned attention mechanism layer to generate the attention feature information set, including: using the above-mentioned attention mechanism layer, performing the following processing steps: setting the single quantity state influencing factor weights whose weights are lower than the target weight values ​​in the above-mentioned single quantity state influencing factor weight set to predetermined values ​​to obtain the set weight set; performing weighted summation processing on the above-mentioned set weight set and the above-mentioned output information set to generate the above-mentioned attention feature information set.

[0014] Optionally, the above-mentioned first historical user portrait is generated through the following steps: obtaining historical user-related information corresponding to the above-mentioned target user; extracting attribute information from the above-mentioned historical user-related information through a pre-trained attribute information extraction model to obtain an attribute information set; and constructing the above-mentioned first historical user portrait based on the above-mentioned attribute information set.

[0015] In a second aspect, some embodiments of the present disclosure provide an information generating device, including: a first acquisition unit, configured to acquire user-related portraits and historical order quantity related information corresponding to a target user; a first generation unit, configured to generate a logistics forecast order quantity within a target time period based on the above-mentioned user-related portraits and the above-mentioned historical order quantity related information, using a pre-trained logistics order quantity generation model, wherein the above-mentioned logistics order quantity generation model is trained based on a target training sample set, the target training samples are training samples whose first logistics platform does not affect the corresponding order quantity changes of users, and the above-mentioned logistics forecast order quantity is a representation of the forecast order quantity without being affected by the above-mentioned first logistics platform; a second acquisition unit, configured to acquire the actual logistics order quantity of the above-mentioned target user within the above-mentioned target time period; a second generation unit, configured to generate order quantity impact information within the above-mentioned target time period based on the above-mentioned actual logistics order quantity and the above-mentioned logistics forecast order quantity, wherein the above-mentioned order quantity impact information represents the order quantity changes under the influence of the above-mentioned first logistics platform.

[0016] Optionally, the user-related portraits include: a first historical user portrait corresponding to the target user, and a second historical user portrait corresponding to a similar user; the historical order quantity related information includes: first historical order quantity change information corresponding to the target user, and second historical order quantity change information corresponding to the similar user; and the first generation unit can be configured to: obtain logistics platform information corresponding to the second logistics platform set, wherein the second logistics platform is a logistics platform that has performed logistics processing for the target user; input the first historical user portrait, the second historical user portrait, the first historical order quantity change information, the second historical order quantity change information and the logistics platform information into the logistics order quantity generation model to generate the logistics predicted order quantity.

[0017] Optionally, the device also includes: obtaining a weight set of single quantity status influencing factors corresponding to the single quantity status influencing factor set for the above-mentioned target time period; obtaining a historical influencing factor data set corresponding to the above-mentioned single quantity status influencing factor set; for each user order quantity change state in at least one user order quantity change state, executing the following first generation step: obtaining a user order quantity change information generation model corresponding to the above-mentioned user order quantity change state; generating user order quantity change information using the above-mentioned user order quantity change information generation model according to the above-mentioned user-related portrait, the above-mentioned historical influencing factor data set and the above-mentioned single quantity status influencing factor weight set; in response to determining that the order quantity change situation represented by the above-mentioned user order quantity change information is consistent with the order quantity change situation corresponding to the above-mentioned user order quantity change state, determining at least one single quantity status influencing factor in the above-mentioned single quantity status influencing factor set whose corresponding size of the corresponding single quantity status influencing factor weight is within the previous target number as at least one single quantity change attribution factor.

[0018] Optionally, the above-mentioned single quantity status influencing factor weight set is generated by the following steps: for each single quantity status influencing factor in the above-mentioned single quantity status influencing factor set, execute the following second generation step: obtain historical influencing factor data corresponding to the above-mentioned single quantity status influencing factor in a historical time period as target historical influencing factor data; input the above-mentioned target historical influencing factor data into the corresponding pre-trained time series neural network model to generate influencing factor feature information; input the above-mentioned influencing factor feature information, the first historical user portrait and the second historical user portrait into the pre-trained single quantity status influencing factor score generation model to generate the single quantity status influencing factor score; normalize the obtained single quantity status influencing factor score set to generate the above-mentioned single quantity status influencing factor weight set.

[0019] Optionally, the above-mentioned user order quantity change information generation model includes: a dense layer, an attention mechanism layer, an attribute feature information generation model and a user order quantity change information generation layer; and the device also includes: inputting the obtained influencing factor feature information set into the above-mentioned dense layer to generate an output information set; inputting the above-mentioned output information set and the above-mentioned order quantity state influencing factor weight set into the above-mentioned attention mechanism layer to generate an attention feature information set; obtaining the user attribute information set corresponding to the above-mentioned target user; using the above-mentioned attribute feature information generation model to generate attribute feature information for the above-mentioned user attribute information set; performing feature information fusion on the above-mentioned attribute feature information and the above-mentioned attention feature information set to generate fused feature information; inputting the above-mentioned fused feature information into the above-mentioned user order quantity change information generation layer to generate the above-mentioned user order quantity change information.

[0020] Optionally, the device also includes: using the above-mentioned attention mechanism layer to perform the following processing steps: setting the weights of the single quantity state influencing factors whose weights are lower than the target weight value in the above-mentioned single quantity state influencing factor weight set to a predetermined value to obtain the set weight set; performing weighted summation processing on the above-mentioned set weight set and the above-mentioned output information set to generate the above-mentioned attention feature information set.

[0021] Optionally, the above-mentioned first historical user portrait is generated through the following steps: obtaining historical user-related information corresponding to the above-mentioned target user; extracting attribute information from the above-mentioned historical user-related information through a pre-trained attribute information extraction model to obtain an attribute information set; and constructing the above-mentioned first historical user portrait based on the above-mentioned attribute information set.

[0022] In a third aspect, some embodiments of the present disclosure provide an electronic device comprising: one or more processors; a storage device on which one or more programs are stored, and when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation manner in the first aspect.

[0023] In a fourth aspect, some embodiments of the present disclosure provide a computer-readable medium having a computer program stored thereon, wherein when the program is executed by a processor, the method described in any implementation manner in the first aspect is implemented.

[0024] In a fifth aspect, some embodiments of the present disclosure provide a computer program product, including a computer program, which implements the method described in any implementation manner in the above-mentioned first aspect when executed by a processor.

[0025] The above-mentioned embodiments of the present disclosure have the following beneficial effects: through the information generation method of some embodiments of the present disclosure, the influence of the first logistics platform on the change of the order quantity of the target user within the target time period can be accurately generated. Specifically, the reason for the inaccurate influence of the relevant order quantity change is that: for the attribution analysis of the change of the logistics order quantity of the user, the existing model often does not consider the influence of the main influencing factors (for example, third-party factors) on the change of the logistics order quantity, resulting in the inaccurate attribution analysis of the change of the logistics order quantity. Based on this, the information generation method of some embodiments of the present disclosure first obtains the user-related portrait and historical order quantity related information corresponding to the target user. Here, the obtained user-related portrait can effectively reflect the various characteristic information of the user corresponding to the target user, so that the subsequent logistics order quantity generation model can obtain various characteristics related to the target user. The historical order quantity related information obtained allows the logistics order quantity generation model to learn more order quantity information related to the above target user. Then, according to the above user-related portrait and the above historical order quantity related information, the pre-trained logistics order quantity generation model can be used to accurately generate the logistics forecast order quantity within the target time period. Among them, the above logistics order quantity generation model is trained based on the target training sample set. The target training samples are training samples in which the first logistics platform does not affect the change in the corresponding order quantity of the user. The above-mentioned logistics forecast order quantity is a representation of the forecast order quantity without being affected by the above-mentioned first logistics platform. Here, through the logistics order quantity generation model, the natural order quantity of the target user in the target time period can be accurately generated without being affected by the first logistics platform. Then, the actual logistics order quantity of the above-mentioned target user in the above-mentioned target time period is obtained. Finally, according to the above-mentioned actual logistics order quantity and the above-mentioned logistics forecast order quantity, the order quantity impact information in the above-mentioned target time period can be accurately generated to determine the degree of influence of the first logistics platform on the change in the corresponding order quantity of the target user, so as to facilitate the subsequent analysis of the change in logistics order quantity. Among them, the above-mentioned order quantity impact information represents the change in order quantity under the influence of the above-mentioned first logistics platform. In summary, through the logistics order quantity generation model trained by the training sample set in which the first logistics platform does not affect the change in the corresponding order quantity of the user, the forecast order quantity without being affected by the above-mentioned first logistics platform can be accurately generated to compare with the actual logistics order quantity to determine the impact of the various services provided by the first logistics platform on the corresponding logistics order quantity of the target user, so as to realize the accurate attribution analysis of the change in logistics order quantity. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and that components and elements are not necessarily drawn to scale.

[0027] Figure 1is a schematic diagram of an application scenario of the information generation method according to some embodiments of the present disclosure;

[0028] Figure 2 are flow charts of some embodiments of the information generation method according to the present disclosure;

[0029] Figure 3 is a schematic diagram of generating a first historical user portrait in some embodiments of the information generating method of the present disclosure;

[0030] Figure 4 are flow charts of other embodiments of the information generating method according to the present disclosure;

[0031] Figure 5 is a schematic diagram of the structure of some embodiments of the information generating device according to the present disclosure;

[0032] Figure 6 It is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. DETAILED DESCRIPTION

[0033] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments set forth herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0034] It should also be noted that, for ease of description, only the parts related to the invention are shown in the drawings. In the absence of conflict, the embodiments and features in the embodiments of the present disclosure can be combined with each other.

[0035] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0036] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0037] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0038] With regard to the collection, storage, and use of user personal information (such as user portraits) involved in this disclosure, before performing the corresponding operations, the relevant organizations or individuals shall fulfill their obligations, including conducting personal information security impact assessments, fulfilling the obligation to inform the personal information subject, and obtaining the authorization and consent of the personal information subject in advance.

[0039] The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0040] Figure 1 It is a schematic diagram of an application scenario of the information generation method according to some embodiments of the present disclosure.

[0041] exist Figure 1 In the application scenario, first, the electronic device 101 can obtain the user-related portrait 103 and historical order quantity related information 104 corresponding to the target user 102. In this application scenario, the target user 102 can be user A. The user-related portrait 103 can be: the user portrait corresponding to user A. The historical order quantity related information 104 can be: the order quantity change of user A in the past 4 months. Then, the electronic device 101 can generate the logistics forecast order quantity 106 in the target time period 107 based on the above user-related portrait 103 and the above historical order quantity related information 104, using the pre-trained logistics order quantity generation model 105. Among them, the above logistics order quantity generation model 105 is trained based on the target training sample set. The target training sample is a training sample in which the first logistics platform 108 does not affect the corresponding order quantity change of the user. The above logistics forecast order quantity 106 is a prediction of the order quantity without being affected by the above first logistics platform 108. In this application scenario, the target time period 107 can be "within 1 month in history". The logistics forecast order quantity 106 can be "1000". The first logistics platform 108 may be logistics platform A. Furthermore, the electronic device 101 may obtain the actual logistics order quantity 109 of the target user 102 within the target time period 107. In this application scenario, the actual logistics order quantity 109 may be "899". Finally, the electronic device 101 may generate order quantity impact information 110 within the target time period 107 based on the actual logistics order quantity 109 and the predicted logistics order quantity 106. Among them, the order quantity impact information 110 represents the change in order quantity under the influence of the first logistics platform 108. In this application scenario, the order quantity impact information 110 may be "the order quantity impact information of user A in the historical 1 month content by logistics platform A: +101".

[0042] It should be noted that the electronic device 101 can be hardware or software. When the electronic device is hardware, it can be implemented as a distributed cluster consisting of multiple servers or terminal devices, or it can be implemented as a single server or a single terminal device. When the electronic device is embodied as software, it can be installed in the hardware devices listed above. It can be implemented as multiple software or software modules for providing distributed services, for example, or it can be implemented as a single software or software module. No specific limitation is made here.

[0043] It should be understood that Figure 1 The number of electronic devices in the embodiment is only for illustration. Any number of electronic devices may be provided according to implementation requirements.

[0044] Continue to refer Figure 2 , shows a process 200 of some embodiments of the information generation method according to the present disclosure. The information generation method comprises the following steps:

[0045] Step 201, obtain the user-related portrait and historical order quantity related information corresponding to the target user.

[0046] In some embodiments, the execution subject of the above information generation method (for example Figure 1 The electronic device 101 shown can obtain the user-related portrait and historical order quantity related information corresponding to the target user through a wired connection method or a wireless connection method. Among them, the above-mentioned target user can be a user for whom order quantity impact information is to be generated for the first logistics platform. The first logistics platform can be a platform that provides a series of logistics services for the target user, that is, the first logistics platform can be a third-party logistics service platform. For example, the logistics service can be but is not limited to at least one of the following: logistics delivery, logistics storage, and logistics transfer. Order quantity impact information can be information that affects the change of logistics order quantity. The user-related portrait can be a user portrait that has a certain correlation with the target user. In practice, the user-related portrait includes: the user portrait corresponding to the target user. In practice, the user portrait can present the user's feature situation under the corresponding feature set. The historical order quantity related information can be information related to the target user and related to the logistics order quantity in the historical time period. For example, the historical order quantity related information can include: the logistics order quantity change information of the target user in the predetermined historical time period. In the logistics scenario, the corresponding target user can be a customer who uses the logistics service provided by the platform.

[0047] Step 202, based on the above-mentioned user-related portraits and the above-mentioned historical order quantity related information, use the pre-trained logistics order quantity generation model to generate the logistics forecast order quantity within the target time period.

[0048] In some embodiments, the execution subject may generate a predicted logistics order quantity within a target time period using a pre-trained logistics order quantity generation model based on the user-related portrait and the historical order quantity related information. The logistics order quantity generation model may be a neural network model that has been trained and used to generate logistics orders in a future predetermined time period. In practice, the logistics order quantity generation model may be a time series neural network model, for example, the logistics order quantity generation model may be a long short-term memory network (LSTM, Long Short-Term Memory) model. The predicted logistics order quantity may be the predicted logistics order quantity of the target user. The logistics order quantity generation model is trained based on a target training sample set. The target training sample is a training sample that does not affect the change in the corresponding order quantity of the user on the first logistics platform. That is, the target training sample is a specific training sample. In practice, a sample that does not have a logistics operation with the first logistics platform may be obtained as a target training sample. The first logistics platform affecting the change in the corresponding order quantity of the user may characterize the change in the order quantity issued by the user under the corresponding logistics service of the first logistics platform. For example, the high-quality service corresponding to the first logistics platform may cause the user to issue more logistics orders. The low-quality service corresponding to the first logistics platform may cause the user to place fewer logistics orders. The above-mentioned logistics forecast order quantity represents the forecast order quantity without being affected by the above-mentioned first logistics platform. That is, the logistics forecast order quantity can be the natural logistics order quantity corresponding to the target user without receiving the corresponding logistics service of the first logistics platform.

[0049] As an example, the above-mentioned execution entity can directly input the above-mentioned user-related portraits and the above-mentioned historical order quantity related information into the logistics order quantity generation model to generate the logistics forecast order quantity within the target time period.

[0050] In some optional implementations of some embodiments, the user-related portraits include: the first historical user portrait corresponding to the target user, and the second historical user portrait corresponding to the similar user. The historical order quantity related information includes: the first historical order quantity change information corresponding to the target user, and the second historical order quantity change information corresponding to the similar user. The first historical user portrait may be the actual user portrait of the target user in a predetermined historical time period. The first historical user portrait may represent various user feature information of the target user in a historical time period. Similar users may be users with at least one similar feature to the target user. For example, similar users may be users with similar logistics scales to the target user. The second historical user portrait may be the actual user portrait of the similar user in a predetermined historical time period. For example, the predetermined historical time period may be within the past 12 months. The first historical order quantity change information may represent the corresponding logistics order quantity change of the target user in the historical time period. The second historical order quantity change information may represent the corresponding logistics order quantity change of similar users in the historical time period.

[0051] Optionally, the above-mentioned method of generating a predicted logistics order quantity within a target time period based on the above-mentioned user-related portrait and the above-mentioned historical order quantity related information by using a pre-trained logistics order quantity generation model may include the following steps:

[0052] The first step is to obtain the logistics platform information corresponding to the second logistics platform set. The second logistics platform is a logistics platform that has performed logistics processing for the target user. That is, the second logistics platform information can be a platform that has historically provided logistics services to the target user.

[0053] In the second step, the first historical user portrait, the second historical user portrait, the first historical order quantity change information, the second historical order quantity change information and the logistics platform information are input into the logistics order quantity generation model to generate the logistics predicted order quantity.

[0054] Optionally, the first historical user portrait is generated by the following steps:

[0055] The first step is to obtain the historical user-related information corresponding to the target user. The historical user-related information can be various unstructured text information related to the target user at historical time. In practice, the historical user-related information can include but is not limited to at least one of the following: text information registered when the user signs the contract, the words and feedback information of visiting the user, and the user encyclopedia information crawled from the Internet. The historical user-related information contains a lot of high-value user attribute information.

[0056] In the second step, attribute information is extracted from the above historical user-related information through a pre-trained attribute information extraction model to obtain an attribute information set. Among them, the attribute information extraction model can be a neural network model for extracting attribute information. In practice, the attribute information extraction model can be a GPT-4o large language model, or a CRF extraction method. The attribute information set can include but is not limited to at least one of the following: main business information, revenue scale information, competitor information, operating platform information, and demand point information. The main business information can be the user's main business type and shipping category, etc. Revenue scale information can include: the revenue scale of the target user, the revenue scale of similar users. Competitor information can characterize the user's order processing on other logistics platforms. The operating platform information can be the logistics platform used by the user. Demand point information can be the demand point that the user is concerned about when shipping (for example, timeliness, price discounts, safety, etc.).

[0057] The third step is to construct the first historical user portrait based on the above attribute information set.

[0058] See also Figure 3 , showing a schematic diagram of generating the first historical user portrait.

[0059] First, the target user's corresponding contract registration information, visit feedback information, and encyclopedia information obtained through crawlers are obtained. Then, the contract registration information, visit feedback information, and encyclopedia information are stored in the knowledge base. Next, the first historical user portrait is generated through the prompt word (prompt) using the GPT-4o large language model.

[0060] Step 203, obtaining the actual logistics order quantity of the above target user within the above target time period.

[0061] In some embodiments, the execution subject may obtain the actual logistics order quantity of the target user in the target time period by wired or wireless means, wherein the actual logistics order quantity may be the real logistics order quantity actually issued by the target user in the target time period.

[0062] Step 204: Generate order quantity impact information within the target time period based on the actual logistics order quantity and the predicted logistics order quantity.

[0063] In some embodiments, the execution entity may generate order quantity impact information within the target time period based on the actual logistics order quantity and the predicted logistics order quantity. The order quantity impact information represents the change in order quantity under the influence of the first logistics platform. In practice, the order quantity impact information may include: impact information representing that the first logistics platform has a positive effect on the logistics order quantity corresponding to the target user, impact information representing that the first logistics platform has a negative effect on the logistics order quantity corresponding to the target user, and impact information representing that the first logistics platform has no effect on the logistics order quantity corresponding to the target user.

[0064] As an example, first, the above-mentioned execution entity can subtract the actual logistics order quantity from the predicted logistics order quantity to obtain a subtraction value. Then, in response to determining that the subtraction value is greater than the first predetermined value, influence information is generated that characterizes that the first logistics platform has a positive effect on the corresponding logistics order quantity of the target user. Next, in response to determining that the subtraction value is less than the second predetermined value, influence information is generated that characterizes that the first logistics platform has a negative effect on the corresponding logistics order quantity of the target user. Among them, the first predetermined value is greater than 0. The second predetermined value is less than 0. Finally, in response to determining that the subtraction value is greater than or equal to the second predetermined value, and less than or equal to the first predetermined value, influence information characterizing that the first logistics platform has no effect on the corresponding logistics order quantity of the target user is generated.

[0065] The above-mentioned embodiments of the present disclosure have the following beneficial effects: through the information generation method of some embodiments of the present disclosure, the influence of the first logistics platform on the change of the order quantity of the target user within the target time period can be accurately generated. Specifically, the reason for the inaccurate influence of the relevant order quantity change is that: for the attribution analysis of the change of the logistics order quantity of the user, the existing model often does not consider the influence of the main influencing factors (for example, third-party factors) on the change of the logistics order quantity, resulting in the inaccurate attribution analysis of the change of the logistics order quantity. Based on this, the information generation method of some embodiments of the present disclosure first obtains the user-related portrait and historical order quantity related information corresponding to the target user. Here, the obtained user-related portrait can effectively reflect the various characteristic information of the user corresponding to the target user, so that the subsequent logistics order quantity generation model can obtain various characteristics related to the target user. The historical order quantity related information obtained allows the logistics order quantity generation model to learn more order quantity information related to the above target user. Then, according to the above user-related portrait and the above historical order quantity related information, the pre-trained logistics order quantity generation model can be used to accurately generate the logistics forecast order quantity within the target time period. Among them, the above logistics order quantity generation model is trained based on the target training sample set. The target training samples are training samples in which the first logistics platform does not affect the change in the corresponding order quantity of the user. The above-mentioned logistics forecast order quantity is a representation of the forecast order quantity without being affected by the above-mentioned first logistics platform. Here, through the logistics order quantity generation model, the natural order quantity of the target user in the target time period can be accurately generated without being affected by the first logistics platform. Then, the actual logistics order quantity of the above-mentioned target user in the above-mentioned target time period is obtained. Finally, according to the above-mentioned actual logistics order quantity and the above-mentioned logistics forecast order quantity, the order quantity impact information in the above-mentioned target time period can be accurately generated to determine the degree of influence of the first logistics platform on the change in the corresponding order quantity of the target user, so as to facilitate the subsequent analysis of the change in logistics order quantity. Among them, the above-mentioned order quantity impact information represents the change in order quantity under the influence of the above-mentioned first logistics platform. In summary, through the logistics order quantity generation model trained by the training sample set in which the first logistics platform does not affect the change in the corresponding order quantity of the user, the forecast order quantity without being affected by the above-mentioned first logistics platform can be accurately generated to compare with the actual logistics order quantity to determine the impact of the various services provided by the first logistics platform on the corresponding logistics order quantity of the target user, so as to realize the accurate attribution analysis of the change in logistics order quantity.

[0066] Further references Figure 4 , shows a process 400 of another embodiment of the information generation method according to the present disclosure. The information generation method comprises the following steps:

[0067] Step 401, obtain the user-related portrait and historical order quantity related information corresponding to the target user.

[0068] Step 402, based on the above-mentioned user-related portraits and the above-mentioned historical order quantity related information, use the pre-trained logistics order quantity generation model to generate the logistics forecast order quantity within the target time period.

[0069] Step 403, obtaining the actual logistics order quantity of the above target user within the above target time period.

[0070] Step 404: Generate order quantity impact information within the target time period based on the actual logistics order quantity and the predicted logistics order quantity.

[0071] In some embodiments, the specific implementation of steps 401-404 and the technical effects thereof can be referred to in Figure 2 The steps 201-204 in the corresponding embodiment are not described in detail here.

[0072] Step 405: Obtain a weight set of order quantity status influencing factors corresponding to the order quantity status influencing factor set for the target time period.

[0073] In some embodiments, the execution subject (for example, an electronic device) may obtain a single quantity status influencing factor weight set corresponding to the single quantity status influencing factor set for the above-mentioned target time period. Among them, there is a one-to-one correspondence between the single quantity status influencing factors in the single quantity status influencing factor set and the single quantity status influencing factor weight set. Single quantity status influencing factors may be factors that affect the occurrence of single quantity status. In practice, the single quantity status influencing factor set may include but is not limited to at least one of the following: customer complaint factors, claims factors, timeliness factors, value discount factors, loss factors, and visit factors. The single quantity status influencing factor weight may be a characterization of the importance of the single quantity status influencing factor to the change in the single quantity status. The single quantity status influencing factor weight may also characterize the importance of the corresponding single quantity status influencing factor in the single quantity status influencing factor set equivalent to the change in the single quantity status. The single quantity status influencing factor weight may be a value between 0 and 1. The higher the value, the higher the importance of the characterization to the change in the single quantity status.

[0074] Here, by considering each order quantity status influencing factor in the order quantity status influencing factor set, the feature weights can be debiased based on each customer feature, so as to subsequently accurately generate user order quantity change information.

[0075] In some optional implementations of some embodiments, the above-mentioned order quantity status influencing factor weight set is generated by the following steps:

[0076] In the first step, for each order quantity status influencing factor in the above order quantity status influencing factor set, the following second generation step is performed:

[0077] Sub-step 1, obtaining historical influencing factor data corresponding to the above-mentioned order quantity status influencing factors in a historical time period as target historical influencing factor data. The historical influencing factor data may be a data sequence of the order quantity status influencing factors in a historical time period. For example, the historical time period may be the nearly 4 months before the target time period.

[0078] Sub-step 2, input the above-mentioned target historical influencing factor data into the corresponding pre-trained time series neural network model to generate influencing factor feature information. The time series neural network model can be a neural network model for extracting time series feature information. For example, the time series neural network model can be a long short-term memory network (LSTM, Long Short-Term Memory) model. The influencing factor feature information can be information in the form of a vector under the time series feature.

[0079] Sub-step 3, input the above-mentioned influencing factor feature information, the first historical user portrait and the second historical user portrait into the pre-trained order quantity status influencing factor score generation model to generate the order quantity status influencing factor score. Among them, the order quantity status influencing factor score generation model can be a neural network model for generating the order quantity status influencing factor score. The order quantity status influencing factor score can be a direct representation of the importance of the order quantity status influencing factor to the order quantity status change. The higher the order quantity status influencing factor score, the higher the degree of influence on the order quantity status change. The order quantity status influencing factor score generation model can be a regression model. For example, the order quantity status influencing factor score generation model can be an Xgboost Regressor regression model.

[0080] In the second step, the obtained score set of order quantity status influencing factors is normalized to generate the above-mentioned weight set of order quantity status influencing factors.

[0081] Step 406, obtaining a historical influencing factor data set corresponding to the above-mentioned order quantity status influencing factor set.

[0082] In some embodiments, the execution subject may obtain a historical influencing factor data set corresponding to the order quantity status influencing factor set. There is a one-to-one correspondence between the order quantity status influencing factors in the order quantity status influencing factor set and the historical influencing factor data in the historical influencing factor data set. The historical influencing factor data may be data corresponding to the order quantity status influencing factors in the corresponding historical time period.

[0083] Step 407: for each user order quantity change state in at least one user order quantity change state, perform the following first generation step:

[0084] Step 4071, obtaining a user order quantity change information generation model corresponding to the above user order quantity change status.

[0085] In some embodiments, the above-mentioned execution entity may obtain a user order change information generation model corresponding to the above-mentioned user order change state. Among them, the user order change information generation model may be a neural network model that generates user order change information. User order change information may be information that characterizes the user order change situation under the corresponding user order state. In practice, the user order change state may be a change state in which the user order changes. For example, the user order change state may be one of the following: an order increase potential state, a decrease state, and a loss state. For example, each user order change state may have a user order change information generation model of a corresponding network structure. The network structure corresponding to each user order change state may also be the same. For example, the user order change information generation model may be a neural network model that can perform classification tasks in deep learning. For example, the user order change information generation model may be a convolutional layer with multiple layers connected in series.

[0086] Step 4072, based on the above-mentioned user-related portrait, the above-mentioned historical influencing factor data set and the above-mentioned order quantity status influencing factor weight set, the above-mentioned user order quantity change information generation model is used to generate user order quantity change information.

[0087] In some embodiments, the execution subject may generate user order quantity change information based on the user-related portrait, the historical influencing factor data set, and the order quantity status influencing factor weight set using the user order quantity change information generation model. The user order quantity change information may be one of the following: change information indicating that the user's corresponding order quantity change is consistent with the user's order quantity change status, and change information indicating that the user's corresponding order quantity change is not consistent with the user's order quantity change status.

[0088] As an example, the execution entity may directly input the user-related portrait, the historical influencing factor data set and the order quantity status influencing factor weight set into the user order quantity change information generation model to generate user order quantity change information.

[0089] In some optional implementations of some embodiments, the above-mentioned user single quantity change information generation model includes: a dense layer, an attention mechanism layer, an attribute feature information generation model and a user single quantity change information generation layer. The attribute feature information generation model can be a neural network layer that generates attribute feature information. In practice, the attribute feature information generation model can include: an Embedding layer and a FNN layer. The user single quantity change information generation layer can be a network layer that outputs user single quantity change information. In practice, the user single quantity change information generation layer can be a binary classification network layer based on a Sigmoid function. For example, the attention mechanism layer can be a network layer based on a multi-head attention mechanism.

[0090] Optionally, the execution subject may generate the user order quantity change information by using the user order quantity change information generation model according to the user-related portrait, the historical influencing factor data set and the order quantity status influencing factor weight set, which may include the following steps:

[0091] In the first step, the obtained feature information set of influencing factors is input into the above dense layer to generate an output information set.

[0092] In the second step, the above output information set and the above single quantity state influencing factor weight set are input into the above attention mechanism layer to generate an attention feature information set.

[0093] The third step is to obtain the user attribute information set corresponding to the above target user. The user attribute information is the attribute content of the corresponding user attribute. The corresponding user attribute set includes: customer level, revenue scale, main business, customer industry, shipping place, receiving place, business platform, demand point, product type, age, and economic entity type.

[0094] The fourth step is to use the attribute feature information generation model to generate attribute feature information for the user attribute information set.

[0095] The fifth step is to fuse the attribute feature information and the attention feature information set to generate fused feature information.

[0096] The sixth step is to input the above fused feature information into the above user order quantity change information generation layer to generate the above user order quantity change information.

[0097] Optionally, the execution subject may input the output information set and the single quantity state influencing factor weight set into the attention mechanism layer to generate an attention feature information set, including the following steps:

[0098] Using the above attention mechanism layer, the following processing steps are performed:

[0099] Sub-step 1: setting the weights of the order quantity status influencing factors whose weights are lower than the target weight value in the above order quantity status influencing factor weight set to a predetermined value, and obtaining the set weight set. The target weight value may be a pre-set weight value. The predetermined value may be 0.

[0100] Sub-step 2, performing weighted summation processing on the above-set weight set and the above-mentioned output information set to generate the above-mentioned attention feature information set.

[0101] Step 4073, in response to determining that the order quantity change situation represented by the above-mentioned user order quantity change information is consistent with the order quantity change situation corresponding to the above-mentioned user order quantity change status, at least one order quantity status influencing factor whose corresponding size of the order quantity status influencing factor weight is within the previous target number is concentrated in the above-mentioned order quantity status influencing factor concentration, and is determined as at least one order quantity change attribution factor.

[0102] In some embodiments, in response to determining that the above-mentioned user single quantity change information represents a single quantity change situation that meets the above-mentioned user single quantity change state corresponding to the single quantity change situation, the above-mentioned execution subject can determine at least one single quantity state influencing factor in the above-mentioned single quantity state influencing factor set whose corresponding single quantity state influencing factor weight is in the front target number as at least one single quantity change attribution factor. The target number can be a preset value. For example, the target number can be a value of 1.

[0103] from Figure 4 It can be seen that Figure 2 Compared with the description of some corresponding embodiments, Figure 4 The process 400 of the information generation method in some corresponding embodiments realizes the debiasing processing of the time series features through the accurate weights corresponding to each feature, and more accurately models the user order quantity status change information under the accurate feature importance between the time series features, so as to more accurately determine the main attribution of the change in the user order quantity.

[0104] Further references Figure 5 As an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of an information generating device, and these device embodiments are Figure 2 Corresponding to the method embodiments shown, the information generating device can be specifically applied to various electronic devices.

[0105] like Figure 5As shown, an information generating device 500 includes: a first acquisition unit 501, a first generation unit 502, a second acquisition unit 503, and a second generation unit 504. The first acquisition unit 501 is configured to acquire the user-related portrait and historical order quantity related information corresponding to the target user; the first generation unit 502 is configured to generate the logistics forecast order quantity within the target time period based on the user-related portrait and the historical order quantity related information, using the pre-trained logistics order quantity generation model, wherein the logistics order quantity generation model is trained based on the target training sample set, the target training sample is the training sample that the first logistics platform does not affect the change of the user's corresponding order quantity, and the logistics forecast order quantity is the predicted order quantity that is characterized without being affected by the first logistics platform; the second acquisition unit 503 is configured to acquire the actual logistics order quantity of the target user within the target time period; the second generation unit 504 is configured to generate the order quantity impact information within the target time period based on the actual logistics order quantity and the logistics forecast order quantity, wherein the order quantity impact information represents the change of the order quantity under the influence of the first logistics platform.

[0106] In some optional implementations of some embodiments, the user-related portraits include: the first historical user portrait corresponding to the target user, and the second historical user portrait corresponding to the similar user; the historical order quantity related information includes: the first historical order quantity change information corresponding to the target user, and the second historical order quantity change information corresponding to the similar user; and the first generation unit 502 can be further configured to: obtain the logistics platform information corresponding to the second logistics platform set, wherein the second logistics platform is a logistics platform that has performed logistics processing for the target user; input the first historical user portrait, the second historical user portrait, the first historical order quantity change information, the second historical order quantity change information and the logistics platform information into the logistics order quantity generation model to generate the logistics predicted order quantity.

[0107] In some optional implementations of some embodiments, the device 500 further includes: a weight acquisition unit, a data acquisition unit, and a step execution unit (not shown in the figure). The weight acquisition unit may be configured to: acquire a single quantity state influencing factor weight set corresponding to the single quantity state influencing factor set for the target time period. The data acquisition unit may be configured to: acquire a historical influencing factor data set corresponding to the single quantity state influencing factor set. The step execution unit may be configured to: for each user single quantity change state in at least one user single quantity change state, perform the following first generation step: acquire a user single quantity change information generation model corresponding to the user single quantity change state; generate user single quantity change information using the user single quantity change information generation model according to the user-related portrait, the historical influencing factor data set, and the single quantity state influencing factor weight set; in response to determining that the single quantity change situation represented by the user single quantity change information meets the single quantity change situation corresponding to the user single quantity change state, determine at least one single quantity state influencing factor in the single quantity state influencing factor set whose corresponding single quantity state influencing factor weight is in the previous target number as at least one single quantity change attribution factor.

[0108] In some optional implementations of some embodiments, the above-mentioned single quantity status influencing factor weight set is generated by the following steps: for each single quantity status influencing factor in the above-mentioned single quantity status influencing factor set, the following second generation step is executed: obtaining the historical influencing factor data corresponding to the above-mentioned single quantity status influencing factor in the historical time period as the target historical influencing factor data; inputting the above-mentioned target historical influencing factor data into the corresponding pre-trained time series neural network model to generate influencing factor feature information; inputting the above-mentioned influencing factor feature information, the first historical user portrait and the second historical user portrait into the pre-trained single quantity status influencing factor score generation model to generate the single quantity status influencing factor score; normalizing the obtained single quantity status influencing factor score set to generate the above-mentioned single quantity status influencing factor weight set.

[0109] In some optional implementations of some embodiments, the above-mentioned user order quantity change information generation model includes: a dense layer, an attention mechanism layer, an attribute feature information generation model and a user order quantity change information generation layer; and the step execution unit can be configured to: input the obtained influencing factor feature information set into the above-mentioned dense layer to generate an output information set; input the above-mentioned output information set and the above-mentioned order quantity status influencing factor weight set into the above-mentioned attention mechanism layer to generate an attention feature information set; obtain the user attribute information set corresponding to the above-mentioned target user; use the above-mentioned attribute feature information generation model to generate attribute feature information for the above-mentioned user attribute information set; fuse the above-mentioned attribute feature information and the above-mentioned attention feature information set to generate fused feature information; input the above-mentioned fused feature information into the above-mentioned user order quantity change information generation layer to generate the above-mentioned user order quantity change information.

[0110] In some optional implementation methods of some embodiments, the step execution unit can be configured to: utilize the above-mentioned attention mechanism layer to execute the following processing steps to set the weights of the single quantity state influencing factors whose weights are lower than the target weight values ​​in the above-mentioned single quantity state influencing factor weight set to predetermined values, and obtain the set weight set; perform weighted summation processing on the above-mentioned set weight set and the above-mentioned output information set to generate the above-mentioned attention feature information set.

[0111] It is understandable that the units recorded in the information generating device 500 are similar to those in the reference Figure 2 Therefore, the operations, features and beneficial effects described above for the method are also applicable to the information generating device 500 and the units included therein, and will not be described in detail here.

[0112] Reference below Figure 6 , which shows an electronic device (eg, Figure 1 Schematic diagram of the structure of the electronic device 101)600. Figure 6 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0113] like Figure 6 As shown, the electronic device 600 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory 602 or a program loaded from a storage device 608 into a random access memory 603. Various programs and data required for the operation of the electronic device 600 are also stored in the random access memory 603. The processing device 601, the read-only memory 602, and the random access memory 603 are connected to each other via a bus 604. An input / output interface 605 is also connected to the bus 604.

[0114] Typically, the following devices may be connected to the input / output interface 605: an input device 606 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 607 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 608 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 609. The communication device 609 may allow the electronic device 600 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 6 The electronic device 600 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead. Figure 6 Each block shown in the figure may represent one device, or may represent multiple devices as required.

[0115] In particular, according to some embodiments of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, some embodiments of the present disclosure include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In some such embodiments, the computer program can be downloaded and installed from a network through a communication device 609, or installed from a storage device 608, or installed from a read-only memory 602. When the computer program is executed by the processing device 601, the above-mentioned functions defined in the method of some embodiments of the present disclosure are executed.

[0116] It should be noted that the computer-readable medium in some embodiments of the present disclosure may be a computer-readable signal medium or a computer-readable storage medium or any combination of the two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In some embodiments of the present disclosure, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In some embodiments of the present disclosure, the computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0117] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0118] The computer-readable medium may be included in the electronic device; or it may exist independently without being installed in the electronic device. The computer-readable medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device: obtains the user-related portrait and historical order quantity related information corresponding to the target user; generates the logistics forecast order quantity within the target time period using the pre-trained logistics order quantity generation model based on the user-related portrait and the historical order quantity related information, wherein the logistics order quantity generation model is trained based on the target training sample set, and the target training sample is a training sample that does not affect the change of the user's corresponding order quantity on the first logistics platform, and the logistics forecast order quantity is a predicted order quantity that is characterized without being affected by the first logistics platform; obtains the actual logistics order quantity of the target user within the target time period; generates order quantity impact information within the target time period based on the actual logistics order quantity and the logistics forecast order quantity, wherein the order quantity impact information represents the change of order quantity under the influence of the first logistics platform.

[0119] Computer program code for performing the operations of some embodiments of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0120] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0121] The units described in some embodiments of the present disclosure may be implemented by software or by hardware. The described units may also be provided in a processor, for example, may be described as: a processor comprising a first acquisition unit, a first generation unit, a second acquisition unit, and a second generation unit. The names of these units do not, in some cases, constitute limitations on the units themselves, for example, the first acquisition unit may also be described as a "unit for acquiring user-related portraits and historical order quantity-related information corresponding to the target user."

[0122] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0123] Some embodiments of the present disclosure further provide a computer program product, including a computer program, which implements any of the above-mentioned information generation methods when executed by a processor.

[0124] The above descriptions are only some preferred embodiments of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by a specific combination of the above-mentioned technical features, but should also cover other technical solutions formed by any combination of the above-mentioned technical features or their equivalent features without departing from the above-mentioned inventive concept. For example, the above-mentioned features are replaced with the technical features with similar functions disclosed in the embodiments of the present disclosure (but not limited to) and the technical solutions formed.

Claims

1. A method for generating information, comprising: Obtain the user-related profile and historical order information corresponding to the target user; According to the user-related portrait and the historical order quantity related information, a pre-trained logistics order quantity generation model is used to generate a logistics forecast order quantity within a target time period, wherein the logistics order quantity generation model is trained based on a target training sample set, the target training sample is a training sample whose first logistics platform does not affect the change of the user's corresponding order quantity, and the logistics forecast order quantity is a forecast order quantity characterized without being affected by the first logistics platform; Obtain the actual logistics order quantity of the target user within the target time period; According to the actual logistics order quantity and the predicted logistics order quantity, order quantity impact information within the target time period is generated, wherein the order quantity impact information represents the change in order quantity under the influence of the first logistics platform.

2. The method according to claim 1, wherein: The user-related portrait includes: a first historical user portrait corresponding to the target user, and a second historical user portrait corresponding to a similar user; the historical order quantity related information includes: first historical order quantity change information corresponding to the target user, and second historical order quantity change information corresponding to the similar user; and The method of generating a predicted logistics order quantity within a target time period based on the user-related portrait and the historical order quantity-related information using a pre-trained logistics order quantity generation model includes: Obtaining logistics platform information corresponding to the second logistics platform set, wherein the second logistics platform is a logistics platform that has performed logistics processing for the target user; The first historical user portrait, the second historical user portrait, the first historical order quantity change information, the second historical order quantity change information and the logistics platform information are input into the logistics order quantity generation model to generate the logistics predicted order quantity.

3. The method according to claim 1, wherein: The method further comprises: Obtaining a weight set of order quantity status influencing factors corresponding to the order quantity status influencing factor set for the target time period; Obtain a historical influencing factor data set corresponding to the order quantity status influencing factor set; For each user order quantity change state in at least one user order quantity change state, the following first generation step is performed: Obtaining a user order quantity change information generation model corresponding to the user order quantity change state; Generate user order quantity change information by using the user order quantity change information generation model according to the user-related portrait, the historical influencing factor data set and the order quantity status influencing factor weight set; In response to determining that the order volume change situation represented by the user's order volume change information is consistent with the order volume change situation corresponding to the user's order volume change status, at least one order volume status influencing factor whose corresponding weight of the order volume status influencing factor concentration is within the front target number is determined as at least one order volume change attribution factor.

4. The method according to claim 3, wherein: The order quantity status influencing factor weight set is generated by the following steps: For each order quantity status influencing factor in the order quantity status influencing factor set, the following second generating step is performed: Acquire historical influencing factor data corresponding to the order quantity status influencing factor in a historical time period as target historical influencing factor data; Inputting the target historical influencing factor data into the corresponding pre-trained time series neural network model to generate influencing factor feature information; Inputting the influencing factor feature information, the first historical user portrait, and the second historical user portrait into a pre-trained order quantity status influencing factor score generation model to generate an order quantity status influencing factor score; The obtained order quantity status influencing factor score set is normalized to generate the order quantity status influencing factor weight set.

5. The method according to claim 4, wherein: The user order change information generation model includes: a dense layer, an attention mechanism layer, an attribute feature information generation model and a user order change information generation layer; and The generating the user order quantity change information by using the user order quantity change information generation model according to the user related portrait, the historical influencing factor data set and the order quantity status influencing factor weight set includes: Inputting the obtained influencing factor feature information set into the dense layer to generate an output information set; Inputting the output information set and the single quantity state influencing factor weight set into the attention mechanism layer to generate an attention feature information set; Obtaining a user attribute information set corresponding to the target user; Using the attribute feature information generation model, generating attribute feature information for the user attribute information set; Fusing the attribute feature information with the attention feature information set to generate fused feature information; The fused feature information is input into the user order quantity change information generation layer to generate the user order quantity change information.

6. The method according to claim 5, wherein: The step of inputting the output information set and the single quantity state influencing factor weight set into the attention mechanism layer to generate an attention feature information set includes: Using the attention mechanism layer, the following processing steps are performed: The weights of the order quantity status influencing factors whose weights are lower than the target weight values ​​in the order quantity status influencing factors weight set are set to predetermined values ​​to obtain a set weight set; The set weight set and the output information set are weightedly summed to generate the attention feature information set.

7. The method according to claim 2, wherein: The first historical user portrait is generated by the following steps: Obtaining historical user related information corresponding to the target user; Extracting attribute information from the historical user-related information through a pre-trained attribute information extraction model to obtain an attribute information set; The first historical user portrait is constructed based on the attribute information set.

8. An information generating device, comprising: A first acquisition unit is configured to acquire a user-related portrait and historical order quantity related information corresponding to a target user; A first generating unit is configured to generate a logistics forecast order quantity within a target time period based on the user-related portrait and the historical order quantity related information using a pre-trained logistics order quantity generation model, wherein the logistics order quantity generation model is trained based on a target training sample set, the target training sample is a training sample whose first logistics platform does not affect the change of the user's corresponding order quantity, and the logistics forecast order quantity is a forecast order quantity characterized without being affected by the first logistics platform; A second acquisition unit is configured to acquire an actual logistics order quantity of the target user within the target time period; The second generating unit is configured to generate order quantity impact information within the target time period according to the actual logistics order quantity and the predicted logistics order quantity, wherein the order quantity impact information represents the change of order quantity under the influence of the first logistics platform.

9. An electronic device, comprising: one or more processors; a storage device having one or more programs stored thereon, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 7.

10. A computer readable medium having a computer program stored thereon, wherein: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

11. A computer program product, comprising a computer program, which, when executed by a processor, implements the method according to any one of claims 1 to 7.

Images